The present invention relates to computing systems, and more specifically, to systems and methods for problem determination in multi-mainframe or hybrid computing environments.
In a typical multi-mainframe computing environment, a mainframe can include multiple servers, For example, the IBM® zEnterprise® system is a hybrid computing environment consisting of classic IBM® mainframe processors running the z/OS® operating system with closely attached commodity processors such as Intel® x86 or IBM® Power® processors which are running housed in one or more IBM® zBX BladeCenter® units and running the Linux® operating system. The zEnterprise® system also houses special versions of IBM® mainframe processors running a version of Linux® called zLinux. The coupling between z/OS® and the blades is accomplished by Ethernet networks using the Transmission Control Protocol/Internet Protocol (TCP/IP) protocol. If any data needs to be transferred from the mainframe to a blade, it must pass over this Ethernet network.
One intended exploiter of these commodity processors is an accelerator such as the IBM® DB2® Analytics Accelerator for query acceleration. Another intended exploiter is for hosting general purpose applications. The goal in most cases is to achieve greater economy in the cost of raw processing, without sacrificing the properties of comprehensive manageability offered in the mainframe environment. The processes on the blades should be experienced by the customer as integrated extensions to the mainframe functionality, not as independent systems with a separate burden of administration. When those processes are managed in this way they can be regarded as agents of the mainframe application(s).
Operation of an agent enabled mainframe includes deployment of software to the blade platform(s) through mainframe software facilities; issuing instructions from mainframe software to the agent, directing it to perform operations; delegating access rights to the agent, for data controlled by mainframe software, for the purpose of performing the directed operations; and the mainframe software receiving results of those operations by the agent. Presently, current systems lack efficient problem determination, and First Time Data Capture, in a system with agents.